A method and device for detecting a drivable area

Abstract

The invention discloses a detection method and device for a drivable area, which uses all contact points between target objects and the ground in the acquired current road image as marked points for marking, and connects all marked contact points to obtain a non-closed polyline , mark the object category corresponding to each target object in the non-closed polyline, and obtain the target road image containing the object category labeling result, input the target road image into the drivable area semantic segmentation model, and obtain the predicted pixel point category of each pixel Information, according to the predicted pixel category information and location information of each pixel, filter and cluster all the marked contact points to obtain different types of contact point sets, based on the contact point set, determine the drivable area in the current road image. When the present invention recognizes the drivable area, only the contact point between the target object and the ground is marked as a marked point, thus saving a lot of marking work and shortening the marking time.

Description

technical field [0001] The present invention relates to the technical field of image processing, and more specifically, to a method and device for detecting a drivable area. Background technique [0002] Drivable area (FreeSpace) detection technology is the key technology of assisted driving system and automatic driving system. In the prior art, the detection method of the drivable area is as follows: the image collected by the vehicle camera is divided into different areas according to different objects, and then the drivable area is identified from the segmented areas. [0003] Since the traditional scheme not only marks the boundary points of the object, but also marks the inner area of ​​the boundary points of the object when performing region segmentation, it not only takes a long time, but also requires complicated follow-up processing to obtain the boundary of the drivable region Information, subsequent processing, such as extracting the edge of the region and classi...

AI Technical Summary

Problems solved by technology

[0003] Since the traditional scheme not only marks the boundary points of the object, but also marks the inner area of ​​the boundary points of the object when performing region segmentation, it not only takes a long time, but also requires complicated follow-up processing to obtain the boundary of the drivable region Information, subsequent processing such as extracting the edge of the region and classifying the overlapping parts of the edges of different regions, etc.

Images